Semi-supervised Semantic Segmentation with Error Localization Network
This addresses a chronic issue in semi-supervised semantic segmentation for computer vision applications, offering a robust solution to improve accuracy with limited labeled data.
The paper tackles the problem of performance degradation in semi-supervised semantic segmentation due to errors in pseudo labels by introducing an error localization network (ELN) that identifies and disregards inaccurate pseudo labels during training. It achieves state-of-the-art results, outperforming all existing methods on PASCAL VOC 2012 and Cityscapes datasets.
This paper studies semi-supervised learning of semantic segmentation, which assumes that only a small portion of training images are labeled and the others remain unlabeled. The unlabeled images are usually assigned pseudo labels to be used in training, which however often causes the risk of performance degradation due to the confirmation bias towards errors on the pseudo labels. We present a novel method that resolves this chronic issue of pseudo labeling. At the heart of our method lies error localization network (ELN), an auxiliary module that takes an image and its segmentation prediction as input and identifies pixels whose pseudo labels are likely to be wrong. ELN enables semi-supervised learning to be robust against inaccurate pseudo labels by disregarding label noises during training and can be naturally integrated with self-training and contrastive learning. Moreover, we introduce a new learning strategy for ELN that simulates plausible and diverse segmentation errors during training of ELN to enhance its generalization. Our method is evaluated on PASCAL VOC 2012 and Cityscapes, where it outperforms all existing methods in every evaluation setting.